LeCun to AI world: share it or shame it
Yann LeCun just dropped a decade-long thesis in one thread: AI, like the printing press and the internet, only works for humanity if it's open. He's backing Mark Zuckerberg's recent warning that concentrating AI power is 'inherently problematic,' and firing back at Dario Amodei's balance-of-power argument by basically saying 'yes, that's me you're talking about.' His pitch: diverse, open foundation models so anyone can build AI reflecting their own values, language and politics, because good AI beats bad AI, not centralized AI beats bad AI.
Nothing says 'trust me' like three tech titans publicly subtweeting each other about who gets to control humanity's future.